Calibration of Piezo-Electric and Electro-Magnetic Engine Indicators Citation Hsu, Chi-Nan (1948) Calibration of Piezo-Electric and Electro-Magnetic Engine Indicators. Engineer's thesis, California Institute of Technology. doi:10.7907/gpmt-vp22. https://resolver.caltech.edu/CaltechTHESIS:05302025-222355880 Abstract No abstract. Item Type: Thesis (Engineer's thesis) Subject Keywords: (Mechanical Engineering) Degree Grantor: California Institute of Technology Division: Engineering and Applied Science Major Option: Mechanical Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Kyropoulos, Peter R. (advisor) Hudson, Donald E. (advisor) Thesis Committee: Unknown, Unknown Defense Date: May 1948 Record Number: CaltechTHESIS:05302025-222355880 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:05302025-222355880 DOI: 10.7907/gpmt-vp22 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17319 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 26 Jun 2025 18:49 Last Modified: 26 Jun 2025 19:14 Thesis Files PDF - Final Version See Usage Policy. 27MB Repository Staff Only: item control page

CALIBR A~.' IO N OF PIE Z0- 1•,LEC'l'RIC AND ELECTn. o -- r. .:i A. mm11 IC ENGI NE I NDIC A'l'OHS 1 Thesis b y Chi-nan Bsu I n Parti a l Fulfillment of the Re quirements f o r the Degree of Mechanical Engineer California Institute o f Technology Pasade na, California Y:ay,1 9 48 Acknowledgmen t The auther wishes to express his deepest indebtedness to Dr. P.R.Kyropoulos who inspired and directed this research. Re also wishes to thank Dr. D.E. Hudson for his guide and advice in preparing this thesis. Special acknowledgment is made to Nr. H.A. Lassen for his co-oo eration in Pefforming the experiments and for his valuable suggestions in analyzing the experimental data and results. Table of Contents Part T:t tle Page I. Summary l II: Introduction 2 I II. Des cri ption of the indicators 5 A. Piezo- ele ctric t y pe 5 B. E le ctro - magnetic t J pe 8 IV. ~ esting eq uipmen t and procedur e 9 v. Re sults a nd discussion 14 VI. Conclusion 28 VII. List of r e f e renc e s 30 VII I. List of figure'3 and t ables ol IX. Append i x 34 1 I. Summary 'rhe object of the present investigation is to study the behavior of two types of electric pressure indicators for use with high speed engines and to establish t he proper calibration technique. The need for the investigation arose from the complete lack of informatlon in li terature concerning the problem . ~hree piezo-electric and three electro-magnetic indicator pickup elements were investi ga ted as to t he:Lt sensitivity up to a pressure of 950 psi. The step-fµnction method was used and is the si gnificant feature of this calibration. 'The signal trace on the oscilloscope screen was recorded photographically. Results are presented in the form of calibration curves of pressure and voltage output vs. oscilloscope deflection and sensitivity vs. pressure . 1~e calibrating techni q ue is outlined in detail and several difficulties li kely to be encountered are discussed in the Ap pendix. The time response of the elements to pressure chang e is discussed and evaluated by means of theoretical compu tations. Th e behavior of t he pickup elements is discussed and is found to be ~n agreement with predi cti on. 2 II. Introduction Engine indicators were at first introduced primarily for taking indicator cards of steam engines and low-speed stationary oil engines from which the indicated power can be computed. After the evolution of high-speed internal combustion engines, the study of t Le cycle of ev e nts becomes another and more important function of the engine indicator. In fact, as the method for ap n roximating the indicated po,,, -er of high -s peed engines by adding tr,e actual work output of the engine to the frictional losses obtained by motoring or driving the engine with an electric dynamometer is simpler and becomes more widel y used, the sole purpdse cf taking indicator cards in high-speed engines is the study of the pressure variation existing in the cylinde r under various op e rating conditions of load, fuel-air ratio, comp ression, i gni tion and valve timing, and with different fuels. The usual mechanical p istpn type engine indicator as used for lar g e stationary slow-speed stea.1,1 and oil engines gives fairly satisfactory indicati on for engine speeds below 300 r.p.m. For the h i gh speeds as encountered in modern high comnression internal combustion engines , this type of indlcator :ts not suited for indlcation purposes b e cause of the inertia of the moving parts which tends to distort the indicator diagram and makes the indicator piston unable to fo 7 low accurately the rapid pressure change taking place inside the cylinder. Furthermore ,the 3 indicator connection increases t lt e compression space ( I appreciably a n d thus infuences both the compression and combustion, es p ecially in s rr,all htgh-con~p ression engine s. Various other devices have been desigr1ed and developed to meet the needs, sucll as optical indicators , samplingvalve indic a tors, balance- p ressure diaphra~n type indicator and electric indicators.(Ref.1,2,~,4,5) '['he electric t y r)e engine indicators which have become popular in research and development work in e n gines have a pre ssure-sensitive ele ment screwed into the cylinder head,which, when acted on by the c ylinder p ressure, produ ces ru 1 elec tric current varying with the pressure. 'Ch is cur :cen t is then flowin g through an amplifier and is recorded on a cathode-ray oscilloscope screen. There are, in g eneral , four types of electric indicators, 1{nown respectively,as t h e resistance, the magnetic(or inductive), the condenser(or capacitive) and the piezo-electric type. (Ref. 6 ). All these indicators make use o f the prop erty of t he pressure-sensitive element to p ro duce an electric signal varying with t he pressure. '.L':he graphic rec o rd t9.ken on the osci l loscope screen of this electric signal is hence also a record o f the pressure inside tb e cylinder. Piezo-electric quart z t yp~ pressure pickups and eletro-magnetic pickup elements are used in the e ngine laboratory of the California Institute. It was found that there were no reliable calibration data from the Was manufacturers of these elements. Besides, it,_deemed 4 necessary to establish a routine method f or calibration in order to insure reliable results from the pressure indicators. 5 III. Descr~ntion of the Indicators A. Piozo-electric type The piezo-electric indicator is especially suited for laboratory and research work to study the cyclic pressure change in high-spe e d engines b y observation of a continuous indication on the screen of a cathode-ray oscilloscope. This equipment, b e cause of the infinitesimal arn-plitude of movement of its parts under pressure and hence its freedom from inertia, is cap able of following and accurately indicating the . pressure variation at even higher speeds than are encountered normallJ in modern engines. It consists of a pressure e lement using certain pressuresensitive crystalline substances such as Hochelle Salts, tourmaline and quartz which posses the peculiar property that, when they are subj e cted to pressure in a particular direction relative to their cr y stal faces, an electric c ha rge appears on these faces. This effect is called the rrPiezo-electric effect 11 • Quartz and tourr:1aline are used when temperatures are high and Hochelle Salt crystals when terrrnera t 1.i.re s are low. 'l'hes e electric charg es, be sides being in exact proportion to the pressure exerted, also vary instantly with change of pressure. The complete set of instrwnents consists of four parts,namely,the pressure element which converts the mechanj_cal nressures into electric impulses, the amplifier by which these electric impulses are magnified, the cathode -ray oscillosc<D ;Je which produces a g raphic record directly of tlrn arr,plifi e d electric impulses and indir e ctl y of the 6 pressure cycle of the cylinder, and the synchronizer which assumes a definite phase relation between the c -~oss sweep of the os cilloscope and the engine c y cle. The indicator being lnv e stigateci ls of the quartz n1ezo - electric type,consisting of a !! Cox tyr) e- 3 '' pre ssure element, 11 type-2l!portable amplifier,cathode-ray oscilloscope and the '' tyDe-6' 11:1.ul ti p le con tac tor for synchronism. The element(Ref . 7),as shown i n Fig .l,consists of two quartz cry stals mount e d in a small c y l indrical c a sing between grounded electrodes,known as the piston and plug. A third electrode ,i nsulated from ground b y the cry stals themse lves , is ins e rted between the two cr-ystals and is the "high" electrical con..r1ec tion of the unit. When a n external nressure is •exerted on the piston(thus compressing the two crystals),an electric char g e will appear on the two faces of eacb crystal. The crystals are so polari ze d t hat these charges are qdditive, the crystals thus op erating in Darellel. '!;'hen no current is drawn from ttJ"::! unit, the volta g e b e tween the center and ground e lectrode s is directly nro po rtional to this ch ~rge and inversely nro po rtio nal to the ca pacit J between t he tw o electrodes. TYe "Cox t yp e- 2 n amplifier is us e d to magnify the small output volta g e of the pre ssur e element to t h e desired amount ot1 t h e oscilloscope scr een . ,'\. ccording to th e m,9nuf ac t"),lr er s' handbook ( He f. 8 ), it has a minimun frequ en.C :\T r esponse of 1 c. p. s., and is abl e to respond faithfully to a single h1pulse , whe t her it b e i n U _e form of frequency as low as 1 c.p.s.,or any complex wave form 7 1,iri thin t h e frequency range of the amplifier, tbereby making it icleally suited f o r the study of infrequent or non-recurring phenomena . It is a three st age re[sistancecnjJac i ta nce coupl e d aud i o -frequenc y ampl1f'ier. rthe first sta g e has two di f fe re n t input i mpedanc es , wh ich are selected by the po sit5-oning o f th e i npu t nselector" s witch. Pos iti on 1-A should be used for all applica ti cns in wh ic h the impedance of t h e si gna l voltage to be ampl if ie d is either r>esi s tive o r i n cluc ti ve and in which t he arnpli tude of the si g na l volta ge is between 0 .1 of 1 m.v. and 300 m.v. Pos ition 1-B is u ser'l for the signal volta r_ e from p iezoelectric a nd capacitive soueces. Posi t ion 2 connects the s i gna l volt age into the second s ta g e of the amp lifi er arid is used for an inpu t vol t a g e of 5 to over 1000 m.v. Be side the above-mentioned input circuits , tl: e amplifier has a lso incor n orated within it a calibra ting volta ge circuit which can be turned on or off. The amount of calibrati ng voltage us e d can be varied b y means c f the volta g e control. The recording or re g ist e ring instrwnent of the arnplified e lectric i mpu l se is th e cathode-ray oscilloscope, while the equipment for synchronizing the hord.zontal ti ming axis of the sco p e to the cranks h aft ang le of the engine is the 11 Cox typ e- 6 11 multiple contactor. For t he co n struction and o pe rati o n o f t he latter, see He f. 9. 8 B. Electro-magnetic typ e 'j'}j e electro- magnetic indicator consists of t Le same pa r t s 8. S t lJe p ie zo- cl ec tric t ype exc e pt tha t t h e pres sure el eP, ent(Hef.10) employs tl:e el e ctro-r~1agnetic :metbod of g enerating e voltage across t h e terminals o f a coil by converting the diaphra gm vibrations i n to v nriation s in ma g netic fhrn through th8 soft steel core · of tb e coi 1. /1.s t h e pressu.re a cting on - the d.L1.9 hra grri :ts cl, ang ed, the mover;1ent or vibr a t:ton of d iaphragm ch ang es the width of the air gap in the ma gn e tic circuit, thereb y c h anging the flux p -::i.ssing through the coil. ':'his J?esml ts in the gener u t t on of induced volta g e in the coil which is directly proportional to t he rate of c h ang e of flux passing through the coil,i . e ., to t he veloci t :y of vibration of dJ.aphr ,J gm unde r pressure chang es. The pressure element,as shown in Fig.2,has two ad ju sting screws to adjus t tbe air gap in the ma g netic circuit.sever al thr e ad e d ho l es are p1'ovided in the ada pter of the pic k up so that cold compressed air may be us e d to,cool t:he coil windings wJ:: en the p ic kup is expo s ed to he 8. t during t=·n g ine testing. 9 I V. Te sting equi pne n t a nd p r ocedure The ~q ui prr.en t us e d f o r c a l ibrating th ese t wo kind s o f p ressure e lements co -.s ists o f o. con v ent i onal dead weight t ester u s ed for calibrating Eo urdon t yp e p re s s ure gaug es and a specially de si g ned qui c k a cting p r e ss u re release valv e for appl y ing and re l ea sing p res s ures on t he e l e me nt. ".;-::_;yc e rine is use c' as t:r,e h y dr au .Lic fluid. up o f 'J'he ge neral set the equi pment is s 11own in F i g . o. Fig. ~ . Ge neral Set-up of t h e t es ting equipment f or calib r ati n~ p ress u re elemen t Va lve stan d B . Dead we i ght t ester Pr e s sure element(electro - magne ti; t Jp e h ere) Os cil loscope E . Amp li f ~e r l't ov i e came ra G. Vlewing hood I-: . ).ui c k R C ting p res sure relea se v al ve I. Tf 8-nifol d A. C. D. F. r:rh e e lemen t mH.ni fold o f t o be calibr a t ed i s scr ewed into t he the v a lv e stand . Bef ore scr ewing the e leme nt in p lace, g l Nc eri ne is bled to el Ln i na t e t >e air trap p ed in the 10 manifold. The element is connected to the input side of the amplifier by means of a shielded cable. 'Th e amplifier gain control should be set a.t an arbi tra:cily fixed position (about 1/3 turn clockwise) such that there will be no noverload i ngrrof the amplif c ier. 'J'his is done by ad,justing the gain control so that when the ~nplifier calibrating . volt ag e is about twic e that corresponding to max imum pressure encount0red,there will be no flattening of the pe .:=iks of the sinusoidal waves on the osciLLosco p e screen. ~his is very important as the overloading of excessive ga i n of ampl:l.fier will,csuse distortion of the pressure diagrams and result in faulty scope d e flections, e specially at hi gh pressur e s or voltages. ', his amplifier g ain positi on shoili.ld be observed throughout t:h e test and further adjustment of the amplitude s of deflections on the scon e screen for dif f erent p ressures and volt a ges is made by means of the sco ~e vertical g ain control. 7he press u re relesse valve is so construc te d that it will g ive a single non - recurring electric sign al set upby the pressure element as the :p ressure is reJ.eased . A step-fundtion wave is thus set up,which is the sign i ficant feature of this cal.ti b ration method . The v alv e assembly consists of a valve,a stand and ~manifold which is connected to the dP ad weight tester through ni p es and a needle valve. ~he oressure rel e ase valve itself is a ball chec k valve consisti n g of three steel balls placed one u pon the other (Fig,4). The balls are first brought int o p ro u er p osition by 11 screwing in the top and slde sc rews. Then the side screw is removed, leaving the balls firmly clamped u pon the valve seat b y means of the top screw. The s:de pressure-relief hole is closed by the bottom ball at th is time. A known we i gh:t is then placed on the dead weight te ster, the needle valve is ope ned and g l yce rin e i s pumpe d through the p ipe line,manifol d an d onto the element, until the weights are lifted in the tester. This means that tbe pressure as recorded on the weights has been applied to the pressure element. 'The weights are then spful, the needle v a lve is cl o sed and the valv e s pring stem i s simul taneo '_1sly pulled out a gainst the spring . The stem is then re le ased and is pushed back by th e spring. 'This throws th e middl e ball out o f the cylinder into the ball reciever. 'The fluid inside t h e manifold, being at hi gh er p res sur e , then quickly lifts the bottom b a ll from its seat,the g l ycerine fl ow s out of the man ifold through the side pressure-relief' h ole. thus a known pressure change can be produced. The resulting electric signal is reco rd ed photographically on t he film. For the piezo-electric pr6 ssure el effient, the electrlc volta g e gene rated is directly proportional to the pre3s ure. For the electro-magnetic pickup, the volta ge induced 1n t he coil is directly proportional to thB rate o f c hang e of magnetic flux passing through t he coil, which, in turn, depends on th e rate of c t ·s.nge of the width of the air gap or the velocity of the diaphragm vibr a tion. The vo-l tage record obtain e d on the scope screen therefor e g ives a measure of the veloci t ;1 rather than the amr• li tude of the 12 diaphragm deflection. From the above consideration, it is seen that the relation between g enerated volta g e and pre ssure in an electro-magnetic pickup depends not only on the amount of p ressure applied but also on the time rate at whic h this p ressure is applied. However, by inserting an integrating circuit beb.l\Je e11J. the amplifier and oscilloscope as shown in Fig. 5, the record obtained from the scope will show the variation of the int e grated volta ge across t l1e condenser and can be u sed as a measure of the amplitude of the d1aphragm vibration • .Since the inertia of the diaphragm is small and can be neglected, the amplitude of the diaphragm deflection can be consider e d as depen d ing on the amount of pressure only,i.e., the time rate of pressure application can be neg l e cted. ~h erefore, the g rap h i c p icture of the integrated volta ge on the scope screen may be used as a measure of the pressu.re applied on the e l emen t. By comparing the scope d e flection corres p onding to a certain pressure ,Ni th that due to a kno wn c a librating voltage, ~he volt ag e corresponding to any unknown pre ssure can be computed. To calibrate t he elements, the procedure is as follov11 s: (1) The equipment is s e t up as shown in Fig.j. ( 2 ) With the amplifier nselecto r" s witch in t he position marked " 1- Cal. 11 , the calibrating volt age is adju s t ed to desired value. The resulting sj_ne wave i3 7e corded by means of a motion picture camera (film moving continuously, shutter removed). 13 (~S) The "Selector" switch is then moved to position 11 1-B r; if a piezo-electric element is used, or to position n1-A 11 if a msgnetic pickup is used. (4) Starting with a pressure of 950 psi, the signal produced by releasing the pressure is recorded on the film. (5) The same procedure is repeated for different pressures. 14 V. Results and Discussion A. Pi e zo-electric typ e Three Cox t yp e- o quartz cr ys tal pressure e lements '. 11 1 ;506 , 4 1286 and ;} 1 2 7 9 were tested. Cox type-2 p ortable amp lifier ( # 1061, Corn.mercial Re search Lab., Inc. ) and c a thode-ray oscilloscope made b y Allen B. Du Mont Lab., Inc.,type 208-B, # 35 7 2 we re used. Severa l s e ts of tests were made for t he el eme nts. Th e results obta i ned for t h e first two sets were inconsist en t a nd are be li eve d to b e in error due to eith er imp ro pe r bleedi ng of air or overloading of t he amplifier. Th ey are shown and discussed in rte tail i n the Ap:;:,end ix. In the t hir d set of test da t ~ , t h e above- men tione d tr oubles are all elimin8. t e d by ttfro ugh bleeding of the t rapp ed a ir i n the pip e lines and by setti n g the amplifiert ga in to a fixed standard p osition so that no overloading could occur. 'P.h e who le p res sure rang e was d ivided into four p arts, ea ch part using a d iffer e nt gain in t he scop e so t h at t h e amp litudes of def l ec ti o n for a ll pre s s ures we re ma de nearl y t h e same.For ea c h ga in s et ting , a certain appropriate calibra ting voltag e was applied . The typical n ictures taken b y th e mov t ng film camer a of the sine waves of the calibrating volt ag e on ti1e sco p e screen and of the ste p -functi on waves p roduced by t he sudden p ressure rele a se a re s hown in Figs.6 and? respectiv e l y . Th e decr e9_ se of amp li tud o s of wav e s for samegain setting as p ressure s are re duced is cl ea rl :· s h own on t h e pie t ure s. The peak -to- peak de flections or amp litudes of the waves as record e d 15 on the film are measured in a comparator with a precision screw of O.ln pitch. A dial is attached to one end of the screw and is graduated i n 100 division~ sb that each division dorresponds to 0.001". The deflections corresponding to a certaih volt age and pressure c an t h us be measured. Th e sensitivity or volta ge output of the element at different pressures can be found as follows: Let x be the defl~ction recorded on the film corresponding to certain calibrating voltage V ( in m. v. ) , y be the deflection corresponding to certain pressure p(in psi), k be sensitivity of the elemen t in m.v./psi, then k=vy/xp. The results of the test are shown in Tables 1,2, and 3 respectively. (All with Urn oscilloscope vertical gain switch at input voltage<..250 v.R. M.S. side.) Table 1. Calibration results for element # 1 306 Scope Ga/n 10 10 10 Co/,'braf/nq Pre55ure{P) Vo!fa9e (V) 950 psi 900 800 Pre5sure Vofra'Je [}erlecfion('j) Oe!lechon(x) 11 o. 315 30 rn. V • 30 30 0.344 0.320 0.287 11.5 700 11.5 600 11. 5 500 20 20 20 0.406 0.345 0.292 o. ;s21 14 14 14 14 400 300 200 100 10 10 10 10 0.424 0.308 0.218 0.092 0.307 0.307 0.307 0.307 30 30 30 75 50 25 () 0.471 0.316 0.112 0.369 0.369 0.369 ,::_, 2 2 ' 0.315 0.015 0.321 0. 321 11 Sens1l v1/J:f K-:: v.Y.;ox j/o/h::,,7e 7enerafed bf/ f'ressure.:: PK. 3. 4 5 rrl.V.,foo psi 32 .8m.v. 3.39 30a5 3.42 27.4 3.62 3.58 3.64 25.5 21. 5 18.2 3.46 ;s . oo 13.8 10.1 7.1 o.O 3 .40 3 .40 2.43 2 .55 1.7 2 0.61 ~~ - ~5 3.55 16 Ta bl e 2 . Cal i bration res u lts f o r el ement /, 1 286 //olfaye Pressure Volfa9e Sens,l/v,i':I' 9ener-a-re d b'1 Def/eclr'on(xJ K::=: //'ll1X J/o/fage(//) Def/echon&, Pres.,ure P~f'K o . 374tt 0 . 442 11 30 m. v . 3 • 7 3. m.lyiio? 3 5.4 m. v. PSI 0 . 42 1 0 . 374 3 . 75 30 33 .7 0 . 368 0. 07 4 3 . 69 30 29 .5 Scope Pressure{?) Co/,'brof/nq Goin 10 10 10 9 50 p si 900 800 11.5 7 00 11. 5 600 1 1. 5 500 20 20 20 o. 4;3 4 0 . 352 0 . 308 0. 327 o . ;52 7 0 . 32 7 3 .7 9 :3 . 59 3 .77 2 6.5 2 1. 5 1 8 .9 14 14 14 14 4 00 300 200 1 00 10 10 10 10 0 . 442 0. 342 0. 222 0 .10 7 0 . 31 8 0 . 318 0. 318 0 . 318 3 . 47 3 . 59 3 . 49 3 . 37 13. 9 1 0.8 6 . 98 3 . 37 30 30 ;)0 75 50 25 2 2 2 0 .4 58 0 . 307 0 .145 0 . 668 O. a 60 O. o60 3 . 39 3 . 41 3 . 22 2 .54 1.72 0 . 81 n2ab le 3 . Ca libr a ti o n res u l t s f or elem en t # 1 '2 7 9 /'resst1-re //o!lac;e • S(:ns1//v,ii Yo/l aye lienerald Scope Pressure (P) Go/n Co/4/m::;h·nq //oll-a.9e (v) fJerlecAm(y) Oe/leclicn (X) 10 10 10 30 m. V • 30 2:>0 0 . 451" 0 . 425 0 . 380 0 . 37 6 11 () . 37 6 0. 37 6 11. 5 700 11.5 600 1 1. 5 500 20 20 20 0. 500 0. 438 0 . 36 2 0 . 385 0 . 385 o. ;335 3 .7 2 3 .7 9 3 . 76 26 .l 22 .7 18 . 8 14 14 14 14 400 :soo 200 1 00 10 10 10 10 0 . 44 1 0 . 34 7 0 . 22 1 0 .1 08 0 . 320 0 . 320 0 . 020 0 . 320 3 .45 ~3 . 62 3 . 46 3 . 68 1 3 .8 10.9 6 . 92 3 . 38 30 30 30 75 50 23 5 2 2 2 0 . 450 0 . 298 0 .1 52 0 . 355 0 . 355 n r 0 . ooo 3 . 38 3 . 36 3 . 42 2 . 54 1. (:)8 9 5 0 p si 900 800 C:-:: K === J/j!/J,,>< by /ress-=J7,:pK ,, 7 9 m.Y./1of> ps, 36 . 0 M. V. 3 .77 c3 .9 3 .7 9 30. 3 0 • 0 . 86 17 'l'he pressure vs. deflection, genera-ted voltage vs. pressure and sensitivity vs. pressure curves fo r all the three p ressure elements are shov1m in Figs.8,9 and 1 0 r e spec ti.vely. The pres sure curves are . all straight lines passing t}1rough the ori.gin, thereby indicating t he linear r ela tion exis ting between the pressure and oscillosc~pe deflection. The gener ated voltage vs. pressure curves are all straight lines pa ssing through the origin,verif-ying the p rediction that the volt age generate d by a piezo-electric quartz element is. di r e ctly proportional to the pressure a cting on the element. Fur the~more ,the three curves are rat h er close t o each ot j er. 'l'he slopes of these lines measure the avera g e sensitiviti e s o f t h e elements ~1306, 4 1 286 and # 1 279 which are 3 . 5 , 3 .75 and 3 .7 5 m.v./100 ps i r espe ctiv e ly. Fig.10 shows that the sensitivity is fairly constant throughout th e wl, ol c pre s sur e range . • lie average sensitivity is 3.5 m.v./100 psi+for ;:~ 1 306 , , 11 hil e tLat for ~1 286 and ! 1279 varies fro m 3.5 to 3.83 m.v./100 psi,the average be i ng about 3 ,7 m.v./1 00 psi. rr'he time respons e o f the e le :il,e n t to pressure cLange can be ev a luated by an exRrnina tion of the wave form in Fig. 7 and also b y a theoretical computation of its natural fr equen cy of vibr a tion. Aft er t he side r1re ssure-relief hok is uncove r e d by the li fting of the bottom ball in t~ e pressure release valve,the pressure insi de the manifold begins to drop. A cert ain ti:oe lag may be present f o r the element to p roduce th e e I ectric signal af t er U ;e pressure cll8.np.:e takes plt~ce. ri'he dis t a nee on tt,e film be tweEn the 18 point where the si gnal begins to be produced and the point where the signal is fully s e t up is found to vary from 1/16 11 8.t high pressures to 0/8n at low pressures. If we assume an average film speed of 3 ft./sec/, then th ese distances cor re spond to 1/576 sec. and 1/96 sec. Ilowe ver,it also takes certain time for t h e release valve to fullJ lift its bottom ball and to complete ly relieve its pressure::. This means that after the electric signal starts to be p roduced, the p ressure conti nues to drop and the signal magnitude continue s to b e b 1i1 i l t up. Some pa rt of t he above-mentioned time duration is hence taken up b y this continuing pressure release. SincA it is expected that the full pressure release will take longer tim~ at low pressures due to the valve action, this exp lains the longer time r ( quired for full set u p of the signal. Based on this consideration,we c a n also conclude that t he time duration o f t he pressure e l emen t to produce si gnal after pressure change takes p lac e is only a oart of t ~1e De riod 1/576 se c. In fact, af~er comp uting its natural fre quency which is found to b e very high,we can predict that this time la g will be very short (much shorter than 1/576 se c.). From F.'ig.l(also Hef.7), .we see that the two quartz cr ysta ls are held betwe en th e steel piston and p lug ~ hlch are both rigid bodied. ~s pres sure is applied on the cr1 stals t h rough the piston, t he cryst 2ls a~e compressed, acting li ke 8. spring, while the piston behaves like a mass attached to th e end of t r,e s p ring. IJ.'he length or thic:Kness o f e11.ch crys t al is f ound from actual measurement to be 0 .2 5'; . 19 The cross -s ec t ional area of the cr:1s:tal and the volwn.<,:; s of' both t he cr ysta ls , .nd piston c a n be measured , and compu t ed . Let p b e the force L, 8.C ting 0:Cl th (0; crys tals, be t h e total length or thickness of the two cr-yst81S= 0 . t50n, I\ be the cross-s e cti ona l = 0 .0 557 S(J . area of the cr y stals in. "· be the modulus of els tici ty of quartz =16 7 ps i, d be the tot a l deflection or deformat i on of the two crystals due to P , t hen P / A. = Ed/I ., or , P/8 = :K = spring cnnstant of the q uartz crJst als -- - u-··• oc:;J,_),: '? x7_ 0 7 /C'' • ,Jre lr ,\ -- 1 • ~.L 1 X 10 6 ,-,_.., ,_ ITJ_, ,::_; .pSl • . Volume of t}, e stee 1 p is ton= C. o :~ 79 cu. in. Volurrie of the two crys tals=0.0 279 cu .in. 3pecifi c ,,re i g}, t o f quar t 2. = 165 lb . / cu.ft. ( Hef . 14) Specifl.ic weight of steel=489 l b./cu . ft.(Ref.14) -= ~: . 46 xl () -4 slug s = m, Mass of the two crystals = O. 0~79 >< lb5/l 729.>C:::'..2 . 2 = 8 . 28 ><-10-5 slug s ::: m2 The natural fr eauency_of t he system Wn~ I<: /( m,+pn. 2 )==jL llxlO"/( 24 . 6 +~;. 76 )===-6. j6Xl0 4 cycles/sec . The high va lue of natural freqUt!:nc;y indicates th e ex c eedingly l o w inertia of t he moving p arts of the e l eme nt, and gi v es a pre diction of quick ti me response to pressure change . Re sides, t he full signal will be set up onl y when 20 the cryst o. l has undergone maximwn compres s ion or deformation b:7 the a ction of t h e pres s ure c hange , i .e.,only when the crys tal has r eached its position at amplitude of vibration. The time re qui red from start to maximum a mp lit-µ de of vibration ca n t h erefore be co nside red as t h e time r espon se of t be element to a fir>st approximation. Tbls is eoual to half the pe riod of vibration or 2 rr/2 1/1/Vh= rr/63600=1/20300 sec.(Note: The vibra tion of the cryst a l is the superposition of its transient term and it s steady term, t~e latter being set u p due to the step-function force or p ressure change). The time for the rapid pressure rise near top dead center a fter igniti o n in internal combustion engin es is found from some p-V diagrams(Fig. 263,p.41:3,Ref.12;Fig. 47,p.91,Ref.13) to be about 5°crank angles at engine speed of '.2000 r.p.m. 'J1his corres p onds to a time duration of 5)(.60/;.:i 6 0x 2000 or 1/2400 sec. If we consider th e time rsponse of t h e cr ys tal eleme nt as CO:\'Uput e d above from its natµral fre~ue ncy to be correct,then it is obvious that the element will be ab le to respond high pressure change s as encountered in modern high speed engines. As the test has been pe rformed with a ll the trap ped air in the pipe lines eli~inated b y p ro per bleediflg and there is no ov er loadi ng of amplifier, t h e result s are suffientl y accurate to warrant conclusions. Some e rrors may be present in s e tting up th e calibrating voltages a nd in measuring deflections on th e film b y the comparator, but t he amount will not be appreciable. In, fact, by t t:trning the comp a rator scr e w in the same direction for some time 21 before taking deflection measurements, and by taking several readings for each measurement, the error can be reduced to a minimum. However,there is some inherent inaccuracy involv e d in the instrument and measurements as discussed in the Appendix, p.34, ~ i g . 18. The a ggregate error that may be possibly involved in the final value of sensitivity can be assumed to be around 10% or± 5;i. 22 B. Electro-magnetic type Thre e electro-magnetic p ressure pickups made by Electro Products Lab., Chicago,111., were tested and calibrated. They are of type 3000A-57, 3 000B- 8 1 and JOOOB-96 r e s pe ctively.The width of air gap in the elements waE fixed b y fj_rst screwing in the pickup in its adapter until it contacts the diaphragm and then turning it backward about 1/2 turn and locking it in place. The elements were calibrated by putting an inte g rating circuit between the amplifier and oscilloscope. For all tests, the amplif :t er g ain was set at the standard position without overloading and the scope go.in was chang ed such that the scop e deflections for various pressures remained essentially the same. Fig.11 shows th e typical pictures taken of the scope deflections for variocs pressures acting on the pickup. The scope trace is continuous a n d .i. s clearly discernible at the ooint of deflection. The deflection is proportional to the amp l:t. tude of diaphragm vibration when a certain pressure change occurs. This deflection decreases as the pressure is lowered(for constant scopB gain setti ng ). The time response of the e le ment to pressure change can be examined in the sa..rne way as for the quartz elenents. The distance on the film between the s. ta rt and completion of set up of the signal varies from l/16r' at high pressures to l/8 11 at l ow pressures,corresQonding to 1/576 a nd 1/::::88 sec. respectively.The natural fr equency of diaphragm vibration is found as follows: 23 From p.432, Ref.11, the natural frequency of v ibration of a ci rcula r plate o f radius r, ma s s per unit area u, is gi v en by: Wn == a JD/u, r 4 where a =1 0 . 21 for clarnped e dges, fundamental mode, D="plate constant"= E t 3 /1 2 (1-u 2 ) E = modul us of 0l estici ty t =th ickne s s of the pla te u = Poissons 1 rat d:o ;:;::: 0.3 For t h e steel diaphragr,1 considered here , t == 0. 0 1 8" for a ll t h e three elemen ts r=0. 2 5 " for 3000A-57 = 0 . ~'81 11 for 3000B - 8 1 and o000 B- 9c. 3 2 ;) =3"-l 0 7 ( 0 . 0 l 8f /12(1-0.3 )=16 lb.-in. ·) en s i ty of the steel pla t e = 489 lb./ c u . ft. (p.549, Ref.14)=0.28 0 lb./c u .in. 4 u , = ( 0 . 2 8 o/ o2 . 2 x 12 ) 0 . 018 = 0 . 1 o2 x 10 - 1 b . s e c ; /cu/in. 4 Wn == 1 0 . 21)1 6/0 . l o 2 xl0- ( 0 . 25 ) 4 =1 8Xl04 cycles/ s e c. for :.S000A- 57 . W.,,==1 0 . 2 1.j 16/0 .1 3 2 x10-4 ( O. 281 )4 =1 4 . 3 xl C4 cycles/ sec. for J0 0 0B - 81 a n d 96 Hence t he natural frequency of t h e diaphr agms is e v en hi gher t h an t h at ofth e quartz crystals. By the same reasoning as for quartz crystals, we c ,n conclud e tha.t these pic k ups have very quick time respons e , and wi ll be able to respond v e r y aui c kly to the rap id pr e ssure ris e i n e n g ine cylinders,. 24 The peculiar wave form in Fig.11 shows the decay of the e lectric charge imparted on the condenser of the lntegrqting circuit bJ the electric im-oulse of the pressure element. A..s tLe pressure change takes place,an electric imnulse :i.s set up by the pickup in ,'lir'ect proportion to the velocity of the diaphragm vibration and is impresse d to the integrating circ ·JJ t, therby imparting electric cru-Jr,ges on the condenser. 1'he condenser for1 .·:3 clcs e d circuits with the oscilloscope and the output side of the amplifier as sl1own inFig. 5. The lealc1g e of this charge through these circuits reduces t}ie voltage across the condenser, therefore score light tr .· ,ce drops. 'l'L1e time :interval requirec for the complete decay determines the wave form tJnd depe,nd.s on the ntime constant 11 of the circuit. Let F be the resistance in tlle clrcu1 t and C be the capacitance of thE; cono.enser, then the tine constant,-is equ8l to HC. For the circuit considered :here, C=0.5 rnfd, R=600,000 ohrns,hence the time constant ->--==HC =O. 5.><.10-6x: 6 xl0 5 = O. o. \. compari sot1 of tbE; decay curves as shown in Fig.186,p.337,Ref.15 and the wave form shown in Pig.11 verifies the above statenent. 3imilarly, the very slow decay of the electric char g e generate d on quartz crystals as shown by the wave forrn in Fig.7 indicates that tLe time constant of the cifcuit there thould be very 'This is quite true because the crystals act like condensers an rl form closed circuit witb the input side of the anplifier. ff'he C [ll!)::lC j_ tance of the crystals is small due to tbe large VG.lue of its thickness but the input impedance of t1,e 8 amnlifier for 1-B circuit as used t:here is very lart,~e(l0 ohms) 25 so that their product willbe lar g e. Following the same procedure as used for tLe calibration of niezo-electric pressure elements,except for t he addition of the integrating circuit and using the 1- A instead of the 1-B circuit,the following results are obtained for the three elements and a re tabulated in Tables 4, 5, and 6. The data are plotted into pressure vs. deflection, sensitivity vs. pressure, f3.nd voltag3 \ C. pressure curves I as shown in Figs.12,13,add 14 respectively. The pressure vs. deflection curves(Figs.1 2 ,13 and 14) show that their relation is linear. The curves are all stright lines passing through t b e origin,indicating that the scope deflection for any gain setting is directly proportional to the pressure applied on the diaphragm. Since this deflection is directly proportional to the inte grated voltage across the condenser of the integrating circuit,which, in turn, is proportional to the amplitude of dianhr11.gm vibration,the amplituc.e oft.be vibration seems to be proportional to the pressure applied to the diaphragm. The three elements have different values of sensi ti vi ty as sho'V'm in F'ig.15, but the modes of variation wit}) respect to pressure for all of them are similar,i.e., sensitivity is larger at pressures between 300 and 700 p si and dro r,s for both higher and l ::wer pressures. 'The maximum value is from 3 to 3.7 m.v./100 psi. 26 Fig.16 indicates the voltage generated b y the p ressure elements as a function of pressure. The curves are not stmght lines,showing that the relation is not linear. This is quite reasonable since the voltage generated is directly proportional to t b e velocit y of the diaphragm vibration which does not depend on the amount of pressure Since there is no mark ed devfuation of the curves and values for all the three elements tested,the results may be considered quite satisfactory. Table 4. Cal:i.bration r e sults for electro-magnetic pickun oOOOB-96 Cal,bra-h'r'l'J Pressure ( P) Pressure Vot+o.1e Vo/t-a<;e (V) oeftecfion('j) Def/ecf ,"on(x) Sensift'v,'tij Volfa.qe lnd=ed JC/_ I<. -- PX 1/;:1 Pressure='Pk ( 1 ) Sco p e gain switch at input voltage< 250 V.R. M.S. side, gain control at 25 0.170:; 0.148Tl ~'). 02 m.vioop.5i 28.7 m.v. 950 psi 25 m.v. 900 25 O.l b3 0.148 3.06 ':: 7.5 850 25 0.17G 0.148 3.50 29.8 800 25 0.160 0.148 3.38 ~:?. 0 (2 ) Sco p e gain switch at input voltage < 250 V.R. M.S. side, gain control at :--so 700 20 0.195 0.154 3.62 25.4 600 3. l:> 8 20 0.170 0.154 22.1 ;';. 67 500 0.154 20 0.141 18.4 ( 3) Scope gain switch at input voltage < 250 V. rL IL S. side, gain control at 50 400 10 0.183 0.129 3.54 14.2 ;300 10 0.129 o.39 0.131 10.2 200 10 0.084 0.129 3.26 G.52 ( 4 ) Scop e gain switch at input voltage< 25 V.R. M.S. gai n control at 60 100 o. 3,:S :3 0.290 2. 3 0 2 75 2 0.245 ~l . C:5 0.290 50 2 0.150 0.2 9 0 2.07 side, 2.30 1.69 1. 0·'1 27 Tabl e 5 . Calibra ti o n results for ele ctr o - mRgne tic pickup 3000B- 81 Pressure ( P) Caftbn:;;l/n9 //off-age(//) p;e-ssure J/0/fate [Jef/ecf/on('I} [)e//echon{x} Jens;l,v,{J;·-Jf<..::: Y!/',P_,)( Vo/fate j'enerq/e,t b'j P === PK ( 1 ) Sc o pe g ain switch at inp ut volt age< ~~ 50 V. R . M.S. sid e , ga in contro l at c·,c;or· 0 . 425n 0 . 4 1 L.l: 11 2 .71 m.v/oo ps,· 25 . 8 m.v. 25 r.i . v. 950 ps i o. 41 4. :::; 3 . 6 o. :.590 2 .7 6 900 25 2·:2 .1 0.:56 5 0 . 414 2 .6 2 GO O 25 ( ( 2 ) :_;cone g ain sv,ri tc h8 t input volta ge< 250 V. R . M.3. side, ga in control at 28 o. ~:) 52 7 00 20 . 4 0 . 398 2 . 91 18 o. ::',52 2 . 95 17.7 6 00 o . :-s4 6 18 500 18 o. 28:5 0 . 352 2 .90 14. 5 ( 3) .S cope ga in switch at input volt age < 250 V . H . :r 1. S . s i de , ga in c on trol at ' .so < .. 0 . 346 0 . 310 2 .79 11. 2 400 10 O. olO 2 .64 7. 9o 10 0 . 245 300 0.31 0 200 0 .1 50 2 . 42 4 . 84 10 ( 4) Scope g ain 2witch at i npu t volt age ~ .25 V. R.fo . S . sid e , ga in c ont rol at :? 6 l CO 2 0 . 368 1. 64 O. o02 1.64 7 ,5 0 . 1 7 /;, 2 0 . 068 1. 25 0 . 93 7 0,114 o . 3tJ8 2 1. 24 0 . 62 0 f>O Table 6 . Calibration results for e lectr o - magne tic pickup ;~:S OOOA - o7 ?res5ure (P) CaiJ brnft"nq flress(,{.re j/01-ra9e Jens/f/v(f~ .K ::=- //51'/;:,.,>( f/o/f-,:z7e fenemfed by p,= ,Pl<... ( 1 ) 3 cope ga in s witch at inpu t vol t a12:e ,_. < 2~)0 ~l . }\ . },~. S . side , g ain control. at 4 0 0 . 234n o . 2oorr m. v. 2 . ~'.5 rn.V}oo~i 21. 4 m.v. 950 psi 0 . 2 ;34 900 25 0. 1 9G 2 .33 21. 0 2 . t;.-S 850 25 0 .1 98 o. 2 ~~4 2 1. 2 r:, oo 25 2 . 46 0 .1 84 0 . 2 34 19 .7 ( ,., \,' S cope ga i n su.ri tc h at i nput vol t ,.1,~ e L.. 2 ::io v . 1-t . r~r . s . s ici e , c: •1in c o ntrol at 60 750 20 0 .26 2 0 . 27 1 19 • I:!_ 2 . 58 o. <J, 4 2 0 . '.2 71 ~:. . 5t 7 00 20 17. 9 600 0 . 205 0 . 271 20 2 . 52 15.1 5 00 20 0 .19 1 0.27 1 14.1 ~- . 82 ( 3 ) :Jco p e up·ain switc h ~l. t inrrn t volt age < 250 V . l~ . Iv, • 3 • side , gain control at 100 400 o. ;?50 0 . 244 ::i . 08 12 l~~. 0 1 ,~ 350 1 0. t:l 0 . 21 2 0.244 2 .9 8 300 12 ~ ; . 84 O.l?o 0 . 244 8 . 52 • +-•~ voltag e < 2b \! . l \ • },~ • S . ( 4) Scope ga in s,,Ji tch at inpu ,3 id e , ga in control 9. t 60 100 0.292 1.73 0 • c:., ;_) ,_) 1.7 ::'"C 75 0 .1 80 2 0.29 2 l. b 4 1. 2~5 50 2 0. 14 i5 1.96 0 <) 0< ) 0. 98 /,/o/1-aqe (Y) Def'/ec/ton(:f. Oer/ecl,a,,(x) ( ) ~ , _ L, ( ' ·,. 1 ,. ,- . ✓. • f..J Cl f__, 28 VI. Conclusion From the expe rime n tal results ob t aine d, we find t h at t h e sensitivity o f niezo- ele ctrlc pressure eleme nt is f a irly const an t throughout the pressure range up to 950 psi, its v a lu e being 3.5 to 3 . 75 m.v . / 1 00 ps i, whi ch is slight l y higher t han the value ( 3 . 0 m.v. / lGG psi ) re p ort ed by the manufact u rer(Ref.7). These values, howev e r, are sub je ctr'dto an error of ± 5% . '1 he general pre diction tha t the elec tr j c 1 voltage gA ner a ted by tte pi ezo-electric effect of quartz crystal e leme nt i s dire ctly p ropo rtional to th e p ressure applied is verified. The linear rel a tion between the scope deflection and pressure i s also estab lished. The volta ge gene r a ted b y e l e ctro- magne tic p ick up under c er tain pressure ~s pr oportional t o the v ~locity of dir~phra grn vi b rati on a nd is not p ro p ort ional to t he pre ssure being app lied. The relation is shown in Fig. 1 6 . 'Lrie s ens itivi t y of t he e l ectro - magn e tic elemen t s is not constant th r oughout the press ure r a n ge tested, but incr ea ses from about 2 . 0 rn.v. / 100 psi at 100 psi to a maximum of about 3 . 0 - 3 .70 m.v./100 p si at pressures ~00 to 700 ps i a nd again drop s as the pressure is further increased(F i g.15 ). The oscilloscope de flection pr oduc ed by the integrat e d volt age is p ro po rtion al to the 8.m.., li tude of diaphr agm vibratio n and to t he p r e s sure applied. The lin e arit y of t he r e lati o n is shown in Fi gs . 12,1 3 and 14. The time response of the elements to pressure change is found to be ver-y quick by considering their high values of natural frequency of vibration.These elements are, therefore, expected tobe suited for high-speed indication applicati o ns. However, this conclusion is based on theoretical computations only, since no definite measurements can be obtained in this respect from the calibr2.ting instruments. VII. List of References (1) .Symposium of Papers on Indicators, Proceeding, Institute of Mechanical Engineers,Vol.1(1923),p.95. (2) Indicators as a Means for Improving aircraft Zngine Performance,by F.L.Prescott,3.A.E.Journal 31.#3,p.361 (3) The Application of an Electro-magnetic Indicator to Internal Combustion Engines by C.3.Draper and J.H. Lauer, K.L.Davis,Journal of leronautical Sciences,Vol.8(194l)p.7 (4) Development and Application of the Cathode-ray Engine Indicator,by ~ .M.Dodds,S.A.E.Journal 39(1936),p.487 • "'h spee d. ,.t, ngine • I nuica -'l • t.or, oy ,_.;. 0. n'1 ay 1 or an d C . ( c)h ) .A. New H _.ig,.~1 • • " S.Draper,Mechanical Engineer,55(1933),p.169 (6) High-spu, d Combustion Engines by P.M.Heldt, (1944,},o.638-647 (7) '}_'he Cox r:~uartz Pressure Elernent,Type 3, Book of Instructions, Commercial Research Lab., Detroit,Michigan (8) The Cox Portable A.mplifier,rl'ype 2,Book of Instructions, Commerci. ·11 Research Lab. ,Detroit,l'i:ichigan (9) The Cox Tf'ultiDlc Contactor,'I'ype 6,Handbook of Instructions, Commercial Research Lab.,Detroit, Eichigan (lO)Instructions for ElectroPickup, Type Series 3000, Electro Products Lab., Chicago, Ill. (ll)~echanical Vibrations,by Den Bartog,p.80,& p.432 (12)Internal Combustion Engines by ~Jennings and Obert (13)Verbrennungsmotoren b;y F.A.F. ;3chmidt (14) Mechanic ·3.1 Engineers' Handbook by L.S. l'ftarks (15)Principle of Electricity by Page and Adams 31 VIII. List of Figures and Tables Figures 1. Cox type-3 quartz crystal pres sure ele:r :1ent, p. 40 2 . Electro-magnetic nressure elei,1ent, P · 40 3. General set-up of testing equipment for calibrating 4. ;J-1.ick-act :t ng pressure releasee valve,P• 41 5. Integrating circuit,p.41 6 . Typical sinusoidal waves of calibrating v oltages on the oscilloscope screen,P-42 7. ~ypical step -function waves due to the electric impulse recorded b y the quartz pressure elernen~~ p.42 8. 0 ressure vs. deflection curves for Cox t J p e- j quartz pressure elements,P• 43 9 . Generated volta.ge vs. presst1.r e curves for Cox type-3 1 quartz pressure elements,P· 44 10. Sensi ti vi t y curves for Cox typr:·- o quartz p re ssul'e e1 emen t21, P.45 11. ~ypical ~ecord of osc illoscone deflections for various pr e ssures acting on elctro - magnetic 9ickup(with integrating circuit)~p.46 12.Pressure vs . deflection curves for electro - ma gnetic pressur B element 30GOB - 96, P· 47 13.Pressure vs. deflection curves for el ~ctro-magnetic p ressure element 3000?-B~P-40 14. 0 ressure vs . defl e ction curves for e lectro - magn etic nres:3ure e lement :'i 0 00.\-57,P•49 15.Sensitivity curves for el e ctro- n a gne t i c p ressure e l em ents p .5() 16. Induced Voltage vs. Pressure Curves for Electro-magnetic Pressure pickups,p.51 17. Pressure vs. Deflection Curves for Cox type~3 Quar tz Pressure Elements (showing the effect of entrapped air for !I 1286).p. 52 . 18. Calibration Volta g e vs. Scope Deflection Curves, showing the errors i n t he experiment,p. 53 19. Sensitivity Curves for Cox type-3 ~uartz PreE 2ure Elements,showing t he faulty result d ue to improper air bleeding,p.54 20. Pressure vs. Deflection Curves for Cox t yp e-3 Quartz Pressure Elements, showing the non-linear relation due to amplifier overloading,P ■ 55 21. Voltage vs. Deflection Curves for Cox type-3 Quartz Pressure Elements, showing t he effect of amplifier overloading,P 56 0 22. Sensitivity Curves for Cox type-o •lu art z Pre ssure El ements , showing low e r sensitivity at high pressure side due to amplifier overloading,P·~7 23. Typical Oscilloscone Deflections f o r Electro-magnetic Pressure Elements( without integrating circui t),p. 58 24. Rotary Selector Valve to Produce Re curring Electric Signal,P ■ 59 Table s 1. Calibrat :i. on r esults for Cox typ e - ::, '' ,J_ urtz ~) ressure c:l. c. ent ,;·1 0 06 , p 15 2 . C~librati on results for Cox t ype - ~ qua rt z p r e s s~r c e l ement ./ 1 286, p.16 3 . Ca lib r a ti on results f o r Cox t ypA - 3 qu2r tz p r e s s u r e 4, . Ca libr a tion r e sults for r0 l e ctro-rnagne tic p ickup 3000B- 9G , P 26 5. CRlibr a ti o ,: results for el e c t:co-rrin.gne tic p r e ssure ,J ickup 3000'S- f3l / p c!i 6 . Calibrati on results for e l e ctro-magne tic p r e ssure pi ckup 30 G0A -57 1 P. c!i IX. A. f\.D;Jend ix Some troubles encounte red i n U :iE: test for piezo-electric nressure elements In the first set of test, the 0 uartz pre ssure el ement s ) 1306 and i.f l286 were calibrated through cmt a pressure range up to 300 ps i only, same s cope g ain setting being u sed for he test. rrhe amplifier ga in position vvas not noticed. 'Ihe data obtaine d for both eleruents are plotted into curves in Figs .17, 1 8 and 19. Cf'he line a r relation be t ween pressure and scoDe deflecti o n for ) 1306 conforms to the or y quite closely whil~ the ~a ta for ~1286 show gres t deviation of the curve from t Le straight line. 't he sensi ti vi ty of el ement /'1306 vari es fro m 2.9 to about 4.0 m.v./lOU psi which is not very dd::fferent from : 0 rn.v./100 psi rE?ported by t he manufactur er , 0 • • while that of !i l286 decr e as e s from ~ . bo to l. 06 m.v./100 psi n.s the p re ssm e is incr en sed . Tbs variation of voltage 0 output or sensitivits of el e ments with r e sp ec t to pressure can b e se e n frmr1 the s e nsi ti vi ty curves in Fig, 19. The cal-ibrati o n voltage vs. scope deflecti on c ur v e s unde r t h ) ..: sme amp li f ier a nd scone ga ln are drawn in thE: same sheet 1 ar•••r son (r.; f or comp c i g . 18) • rnh 1 e a•1 r.o 1erenc e 1 • in deflection for the same calibr8.ting volta ge shows the e rr ors involved in the i nstrun, ents 8nd in d eflect :1.on ireasurements. Jf' therf; is no such err o rs, the t wo curv e s sho ul d coinci rJ.e . 'The spac e between t he two curves therefore is a measur e of t h e e rrors 1::-re sent :l n t he f' xJ,eri ruc:n t. Tak ing, e.g. , t r:,e deflection a t 7 m. v . whi ch i s O . 2 ~n !' for on e curv e and is (.i . 3~9 I/ fer the other. r;'h e percent error is therefor e O. 039-0 . ?,CJ7 /0.-339 or 1 '7 . 4 ~:/ which is rather lRr ,Z: P . 'J'he av t'; rage a g gre g a t e error of the exn eriment can be taken as around 10%. Since the two calibrating voltage vs. scope deflection curves are all straight lines, the r e is no overloading of the amplifier. ''.·be departure of the pressure vs. scope d e flection curve for element fi 1 286 from the theoretical linear r e lati on can therefore not be at t ributed to tl Je amp lifier overloading, but must hqve been cause d b y the pre s ence of air in the in-stru.ment. 'The poor curves are presented h e re to ernphEi_ size tbe need for care i n assemblin g anci bleedl ng of the calibrating s y s tem. Since the high press~re portion of e n g ine p er f ormance is more irnport 'lnt and the maximum combustion p res ~ure i n th e engine cylinders under normal operating conditions is below 1800 psi, the s e cond set of t e st was made for el eme nts i;~l 2 79 and ,f lo06 with the calibrating nressure rang e up to 9b0 psi.Same scope gain was usb ; t h roughout the test and the ar::p lifier g ain positi o n is not not : ced. The d a ta obtained are drawn in Fi g s.20,21 and22. All the p res s ure curves as shown in Flg. 20 are not lin e ar . This , together wiV ~ the s e nsitivity curves,shows that the sensitivity dro p s as the pressure incr e as es. The se n s itivity is in all cases mu ch l ower than 3 . 0 m.v./1 00 psi; the value reported by the ma~~f a ctur e r. The reason for the decr e ase in se n sitivity at high pressures and the depa -rture of the p ressure-deflect io n curves fr rnn the line&r relation may be seen from the calibrating voltage curves in 36 "'ig. 21. -: -, ere the two curv e s are not strai g:b t lines as ant~cipated but becomes stee n er as voltage is increased. Th e p ortions of the curves b e low 8 m.v. are straight lines but t he upper porti on s are curved. Since tl. e amr<Lif:!i.er gain us e rt jn : h e tPs~ tas not been checked for overloading,this pe culiar deviati on may be at t ributed to the overloading of the amplifier, as evj_rrlenced b :; t.he curvature of the voltage curves at high volta ge s. noverloadi ng " caus es disto,rtion of the press·c1.re diagram anl-'l flattens t he peaks of the sine waves, U:ereb y reducing the apparent am:::1li tudes or deflection, especially at h'tgh voltage s or :r; ressures. 1:'his results in faulty sco De deflection Rnd incorrect indicati on of sensi ti vi ty. 37 B. Testing electro-magnetic pickups without integrating circuit If the electro-magnetic pickup elements were calibrated without the integrating circuit, the resulting nictures of t he scope deflections are in the form as shown in Fig,23. These deflections are directly proportional to the velocity of diaphr11gin vibration which depends not only on the amount of pressure apT)lied but also on the; time rate at which this pressure is applied. rr'h e SCOT;e tr a ce as shown on the film cannot be sehn clearly 8.t the point where deflection takes place. As the pressure is suddenly released, the sco -e light trace is first deflected upward and then downward. 'T'he difference between the lowest trace and its equilibrium position becomes smaller as the pressure change ls reduced. At about 150 psi, the trace becomes clearly discernible. The deflection~ of the picture therefore cannot be measured and are useless in evaluating the voltage output or sensitivity of th e element because they are not a measure of pressure only. 38 C. Some sugp;e s tio ns as to the high - ter;ipera ture Calibrat ion of t he auartz o ressure e lements The p ress u re ele1;1ents when scr (-;n·e d into cylinder lie ad and nut in to ac tua l us e are exr;o se d to t ile high tempe ratu res with in the c y li nder . :"o r t h 0 el ec tro-m'l gnct:i. c nj_ ckup , ho J.es a re provided in the adapter s o tha t compressed air ca n be appJ.ied f or coolin g of t he windings and diap hragm. For the p i e zo - ele ctric p ress u r e 8 l ernents, no s u c h co o ling devices are pr ovided so th a t the quartz crys t a ls are subject to the heat of t he conbustion gas in t he c y li nde r. Acc o rding to manufac t-;:rer s I s pe c i :.":i.c a ti on ( Hef . 7 .' , t he sensi ti v i t y of t he Cox ty p e - 3 quartz cryst a l p ress u re elemsnt is unaf fe cted b~ ~ t ffinpe ratu re ilp to J50 °C.( 6 62 ° F).Above this t e mpera tur e , 0 th e qua rtz sensitivity drops r ap idly ,until a t ab out 570 C. (10 !30 °F. ), it is :::e r 0. -1. t thi s temperature, t he q uartz chang es fr om Al pha to Be ta qu 8. rt :?.. r1o wev e r, W)On cooli ng , the quart z. returns to its previ o-i.-1.s form and t Le element r ega ins it s n ormal charac te r:tstics. F r om actual mea suremen ts, t he usual cylinde r head temperature ar e around 6 50°t o 800°F. Hence t he sensitivity of t he quar tz p ressure element when useti in t he en g ine c y lind 0 r will be low e r than its low temperat u re value. Thus, t he oscilloscope record of t h e p ressure v a rlati o n inside t he cylinder gi ves only a qualitat:tve i d e a and c anno t be interpreted quant:i ta ti ve l y unl e ss t he high tempe rature sensi tivi t ~; v ~~_l ue o f t Lo elemen t is k nown . 39 For t h e test performed in t his res ~ ~roh, onl) room tem~erature sensitivity VRlu es of t he ele ments a r e obtained, using the singl e impulse ste T) -function rae thod. In order to conform to actua l eng ine o pe rations mo r e closely, a rotary selector valve should b F-o used so that tlrn pressure eleLtent wlllbe acted on b J high press~re com0re s ~ed air and atmos p heric air alt e rnately wi t h in one r e v 8 luti on of t he v~lve.(FIG· 24 ) \ recurring electric signal willbe generated b y the n ressure element in s t e ad of a single si gna l. The el0rnent is connected to an oscillosco pe t}1rough t he amplifier in tl1e :I s ,J_al v_,ay . By varying t he pressure of t he COJ,'npressed air acting on t h e el ,. -,rnen t and :re co~d ing t Le sco pe~ deflection correspondingly, t h e u sual p ressure vs. deflection curves and sensitivity valu ~ s can be obtained b y means of t he calibrating voltage circu~t as stated in this p~per . For Uie h :1-gh temperature calibration, t.b e same rotary sel e ctor valve driven by a V8 ri ab le-s peed motor can be used, but a heating coil should be ir1stalled ar 0 und t t e ele~en t s o that the lat t er will be heated to any high temnera ture 1_ s re qui red for tl-:e cali bra ti on. rl'he temperature can b e vari e d by m~ans of a resistanc e box plac e d in t he circuit and can b~ measured by a the rmocoup le inserted into q hole in the adapter tnto which th e element is screwed. yigur c s 40 LocJ<nuf Pressure M✓-ca p/u • Screw JVasher.·r---1-_L ~·--..~~=.....!.~ - - -- c- - - - £/eelrode Oua,-lz Cryslol Aressure Elemenf Adapfer Gop Adjusl/n<J Screw coofi,,., 9 Holes £/ecfro-ma~nel/c Pressure •Bc/4up 41 CCI~--rofl Screw Eiec/rdmtiJnel,-c /Jlc,kup co;/ Amp!,l.'e I ~ /npuf lmfea'11t1 Qsc.,//OS"COfe Amp/J)er Ou.1'17ur !m;edl11ce soo,ooo.f2- Conde o.5 M-rd r resisfance 2,0,00, 000..12 i (Belween Arnp/,lr'er and Osol/oscope) . £( 42. 10 m.v. ,S co::ie ga:in 1 4- Pi g . 6. 20 m.v. -S cope grdn 11. S Scope ga•n 10 T7 •i cai sinuso:l.dal waves of cali bra t:ton voltages on t he oscilloscoue screen(Sco pe gain swi tc h at innu t vol tap:;e < -- [)0 V. R. M • .S . st_deJ. 800 psi Sco1~,e ga in 10 90G psi ,' :>co p e gain 10 950 psi Sco rJe gain 10 500 psi Scope ga i n 11. t) 600 psi Sc.o pe g ain 11. 5 700 psi ;:;c o p e t uin 11. 5 100 psi 'Jc o pe gain 1,J 200 psi Scope g ain 1 4 25 p si )co p e gain :SO c>OO p si Sc o pe gain 1 4 50 psi '3co n e gain :z; c Ll OO psi Sco p e gain 1 4 7 !) psi Sco :Js gain :~o Fi g .?. 'Py1yi cal st ep -fun eti o,1 waves d ,ie to t he electri c Inn ulse reco rd ed b7 the quar tz pre s sure element. -500 ps i 600 psi 70 0 ps i 800 psi 200 psi 3 00 psi 400 p si 50 p si 7 5 psi 1 00 ps i 9 00psi F i g .11. Typic a l record of scope d eflecti on s for various pres sures a cti ng on e l ec tromagneti c p ressure e lemen ts ( with int egr a ting circu i t). Scope Ga in: 25 for 900 , BOC psi; 28 f o r 7 00 , 600 , 500 p si; 50 for 400 , jOO , 2 00 psi. ( Scope ga i n s witc h a t inp ut voltag e < 250 V. R. M.S. side. ) 26 f o r 1 00 , ?S , 50 psi . ( Scope gain switch at input volta ge < ;~: 5 V.R. M. S. side ). ,. - -- ·-~ - - ,--_-- ·_ ·- I - _- . • _i _ - -, r 49 52 .i: i :,Tlq7·.-~r1;,: i ::tu ,·-:~ I ' -- .••. ', -· - ' J -- • '-, -·- · - - - ~ --- T ·--· '1 ' ! _i • t:Jtf :H ·' I ,,, 7 ' ' •~ '''·· , /. ' _I l ~ !~ t 1.St ... ' '.Eiot~ ~nm:~~gra~~,,~~;J:•F~~m~~•••i •~ HFi·,, ~~ tffiili'.'.::: -, • •;::. .. T ) ... l .•• I :, .- • • -· i7t7lJ[JJ:ir . • l ' ' - lI I I' • It - ~~-- 1-·--r -:- - -;--={ . ±I' -1 _ __f _'.__,j--,- J-:TJ j - -~.: - ;--· 1-·}---l .J .J._ - ...... - - 1 _ _ l._ _ _ _ 1.. ' _u ' ·1 ' ll •-' -•' I D ----+-4--jf-·::. H~·-- - - ·--·I-· , - L,, -1-,· 1 : : : ' .: •': ! '; i -• I'· l I . j ! ' I •·. . :. . - _;_;_;,L::.J.-.:.. .:..:...1.-I i I I ·: '.- ·-.• ,I ! •r , ' ,. ·---r -···- _, ___1___ .-4w...j~-1---, :I - -'- t ___ l - ; ---L-----,· , jJ •I..:-t:!:...::__'1 :i..;':;)i~;'. 1 ,. • ~' :P~ 1 -, -:+ ' T i"' -j ' l+t·- •··;.,; "" ---:---r:-: I _ _ __ _ _ r .!: I .. ·i·: !f 'd ~ "' - 4 I ' ' 1;1: p · I . ~~~-~, :tl +Ii _ - --i • 1 • • !.)J ::!- I I'-~ . I i----- L__ l __:_ ___ _l _ 1- _ !' •' +--·--~!-:-----·---t·· --~-1--+- -- ;c. " !.ct:R J:l · , ' I I · : - -.'-~I-, :.:. . . . :. i• i II , , . ·, -·-_,i-:T~J&:.ii 2.' · I->- , - ..: ,fr' -- ' ) . ...L _j_ !~··· ! I; 1-·_ " '.., : I - -~r.3~ •I l -·~-lI .:· ' .!..~-+u- ', • ,._,' ' .:,,.:,:: - • • ' ' • ~Ifll~~t::!U•·-,:+,/ L' 1 - ~~~; r~~iit µi! . .'. .. ''.i. '"· "'- .,,.. ,. _., •"' ' , .,. , , ~~"~- ~--··'-~ '-" --- ..' , , 1 - - -, 1 , 1 , - ,-~ttt.· ·'11 , '' : • , . .J I '' l~~·:;: !'. tH ! • .. , ~ ·;_r l '. .. L. ~. ' 'I • ·] 1- 1 •\', .!..l:: '-l _ _F:'...:. " ~ ~Ji~~ 'l ~4-' t • l~ ·-~ •7.-...~ --1 ·1..• ~~- c •-1·.,. l ,- , tt' ,, • .. ! - f - - I -~~-1I2Tl ... I' ' 1 1 _, , ---~-__,~ - , ~ ···.I w,·,tE] ,,.. . ~ • ·• ••• - -" ' ··--- ... , , ·-- - -- - · --- --cl; ; ) '""" ""~ ; " : --· i- : ,..,~ ~"4!-,--,--!-f-'-''-,-;-i,-,+-'--r'+h+h..L/ll-i-'-'-!-/---!-14+;/,~ p:-,- ,,.,._ _ __., ~;t - I • ::7 1 _I -~ -i~-:l I Ile ... , ' 1 Ll '] ti I j - - - ~ :.._• · ~ - • ,_ I r••i r__ rt ..2.., __•___ , 1 1I , • ! : ;;., I'"' c, , 11 , __ _ ~ -! ' ---t ! • • • , .. ' .. 'f -- ' Cl. ,..• .... I - I - •• .. 'J , _, , .. 1 . ,. _ _- __ L_ I ::TI L_ - : ~::: ' -- ~ _ ~: ~ ~-_:, '"I •\ ~ _ . ., • 1: :. I : - - ~~ ~-~ ~.:-1 1 ~I f- I 4" . • ! .:.:... 't< .., · I I J_,:._L_.~-· ~ ~ J' -~- J • • 1 I .. I r -~ : ~-~J•_ .. ! ~ ~ • •..;-:: -,, ~ -M~~;~~-.rt-·-~11:~~~f~-:.;;~/ ~L~ H +H++-+H--+t-4-h f:h' , ~1~.IT.tl 1 .. "'·' . _, r....,_'~--t'-.'.. . [ill",.,·' iil.1 1,' f ' ·"\- 1,·: 1~.. ·- H+i-+r ----h--: ' . , ... r t1 1 ,- ~ J ' I, ·, J :~;-~~~1:\::-~ t} •• • I, '. _J_ I 'i:~:.c:1r :'.f~-J .11 1 · 1·• ~• . ,,., -' ,. r , .., / · 1 ,-4,4-,--4+4'--I r; I~ · . . ,I.. I' · J _J ,.. ,~i.J f \ 1.. • .. , r, t '•·· FTT]':T: -T:1$.- I I 1... ' ~ 1 (:", • ~ --:; ; ; '}1 ITTiTtlf d\ )i: ; r~,I ~t r--m·~----'I'" ,~ - r = - w r " E ' .. ,.. "S"" "" I'"~=t.c . .~l~- -. '~ 1 'I· .. I - · - --- ·-- ~ - '"- - -·--·'"'- ____ ....., ~ ·1 - ' ;, 1 j ir' " · , r''• ·- · 1 • !> · · · · ' I, P" . • • 1 ,I -:!.J-.- ' --=- [ / J• ~.' , __ $ -.. ,., .j '>I"-'"~. ,.. • :..J /t~l_:_:._1-1 1 1 1 r -'- 1 .r~i-i:~1=*!' c I IN:f tAi . tJ :t _";,tFJ-J ~ _: _~:- f~)~~~ r-r~~~t :~·-t ~-~-+:-~ • r .. - "j _'.;~ift~j{) ~:~>~t ') __~ ~-:lfl~•\L~~t{~-;~;- -=-- _:_· -~:. ~~~- ~:J;1~; I' T· .Lf , ,' i-~- r --~, .. a;+'-, ; -iJ - _J.... ,-;-,- .,.,.r- ··H·• -I-!.., •· -'T j""T, l' 1 · : :: • -I . ,I .. • :~ •.: . ; , J". I !"':::-:- .:.: ·-~ :Y;m1;r;:1:t4hi~~ ~,'.~/~~ t '¥ft -til •"I J- " 1}-~V'~i - _,, ;;·,~,,.:, ~;~\]! ••,,c~'.~f ~ d.::: , ~-,,r .! ldi·. - .! .. , 1 1-+ .1.... l ... f+:' ''~','. _., : :. I ,· -:-.: ··, 1-.~: -~~l :.:. :++f~•'-;.j t/'::c. ' ', f ,·:·i ·Lt ,, ,,' i -~ ·~1~',--1 ,.·.: ;..:·'. ': 1 --~ 9 ·-- ~, I ·.l l ·"· .... , .... . ...:..~-=+=-- • -! _ - ·-::....:... •· j·~ ____ __ _ ... .. . . ....->Etf)! 1;?1\ :tr:~:;+' I c4._: ; " "f'""'-l·t ,.Ir .... , .;f_ ~--srt·; . J . ,·1 ,, ' . - .-·--:--r---: :-:-·:- j~?'~;rt~~~-;.;;- .:7ffi#t1 7?:+t+:;_' .; -~~;: ~--~. -~:i•1·~l .. ;·_ ~-: ~i -=-;:j1·---f[ ~ .-:~-~---~ ll . :--+-: • I. ::r.1:+-'-'-- • ~ ~ · ri t , . 1 1:... . :~l ·:J-:~ ~ -~~-r ~~~ !-~ t-~~~ ~: '. . • ~~ -~~!_: __ : ~~ 7-::.i ;~~ -~~~~ ~~ - j. ,.- j-:·-·· _..:. : ·: J ·. .• .l1• : l ,,;~~f 't, klr• j .., T ' i\ ·'1·cl·-l1r'1f-1 ,, l l" I ~! t -JJ# ,f . : . _ _ L _ ~ ~ ~-r• r,- ,~ . . -- · - -· -·•-'--• 1-r-1 -l•~•,1 ~ ~1Li: 1illi_ 1. , •r,* ... . •' • •m·'-•+-1µ.; ur' ,, "-• b -,- 1<10· .eJ..r::.l-r3:]} nT,...e1 n,- J:L, .. j •• , · r •----- '-,'-' ., :. -_·--:.;::·~;;_ :. : - - ~15~_:+ .1:_ ~~-:?:i_-;_~:t_t_::_1_ m !.--~ -•_;-~. r;~£'~_:-i_~ ,Tr~1 1L;.r · '. :::-rrr:~-, 1~·-1-~- ~:-~~ -::~:f' ·: · ·TI.7 : ·- -, 1 ·.·,' ...... . 1.,, ... ...,r t;.6-'·"• -- , .. i.,-1-,. , , . . . . . . · ·· •-~· + ... I _-_:_-i 1 ~~ r.. :-) . , . . (:) I ' . . , 1 •. , 1 , •· -, _ ~ :-1 __ _i__.: 1 .. • -..:....:.J • • ., _ __ _ __J1 · " - · , _ _. _ - _, J -~~ • • -c_ : , . : .- - ,_ ' .:-· . - • r-; r:_: ~-- -~'--~- --. -➔ ~~i-' - ··-- - ' _:_ I ; _:· i - ,jsi' •• _.._. -· ::::i.====_. ~ - -.....-+-'-'-+ -'' " " ' · --,+!~ >· :-·' t+,---,r.+'-;--1¥ -+= .t H-C.'4-1 i'. -"-hA~---,..!.++-.!+-'-'--4-.:.,...C!, :;:;:~.;.x.+;.:.:::;..i:i:;:1:14~q= :!..j.:= + .:..c+:-::..+n=i..q I • 1 -,~_.;..- -'--'-!+.!.+J.lJ.-;.+.!.++'.!.4- ;..cc+,-..!.+:.c+-!-+ \-'·'-+~- -x--+--n~ ;-l>"t:-r-t-~tt-...,...r~-t-, H---,:-:++t-,.;+,..W.-4-,~~ch'-'-++!-'-rHH-"-"--+l-'-'..,+,-,.. ,-'L ~.;+_r.,.~++: ~:,: ::.,._::+~ ·+t,....,--+-~-t+.._,..·.---t ·j-rr,-f--.,.;--j-+"tii f -,-'-!-#-f-,+t-,..W..Y..,..,.C..+- :::::::~i::,:::::::,+i+,-'+J-t'-'- i, -. ,.: :!: :::::::::::~~ ::.+.!..\.!4-!-l.!.:..c;H-:.+µ .:.+1.µ.4-1-.µ I ·= I: :: -- . • +=r- -• ...... ~ - •-· - ~.:--1-:-' ·.-,!_:;.:-4:-- •· : :~?~,; f'i_i ~< , ~1i!j;~ .C:.J..1.,..;.i++!-4-:-:+-!-'-!-'---'+-'-'-l-''-+ •• : ;. L'. .,- '.:J....__._:_:_ I- ;.: 1_ ;;: '._~(: ,. : i·:- • .-,.~ • • :·..•:.; I 57 '.-S 50 p s i 8 50 psi 900 psi 950 psi 4 5 0 psi 550 psi ii50 ps i 1. 00 psi 1 50 psi 2 50 p si Fig . 2 3. Tyo i cal oscilloscon e d eflections for elec tr omagne tic n ressure e l ements( Without int egra t ing circuit) . 59 ' I I ' I